Continuously Evolving and Interactive Disguised Face Identification (DFI) with Facial Key Points using ScatterNet Hybrid Deep Learning (SHDL) Network

Abstract

Disguised Face Identification (DFI) system and method for identifying multiple individuals with disguised faces in uncontrolled environments / scenarios is provided. The Disguised Face Identification (DFI) system and method includes detecting facial landmarks / facial key-points and performing face identification using the ScatterNet Hybrid Deep Learning (SHDL) Network. The system also can be evolved, after deployment, by the user as it provides one with an ability to add new faces to a known face database which are identified by the system thereafter. Further includes two facial disguise (FG) datasets, the datasets are simple facial disguise (FG) datasets and complex facial disguise (FG) datasets for training the deep convolutional networks.

Description

CROSS REFERENCE TO RELATED APPLICATIONS[0001]This application claims priority on U.S. Provisional Patent Application No. 62 / 898,528, entitled “Continuously Evolving and Interactive Disguised Face Identification (DFI) with Facial Key Points using ScatterNet Hybrid Deep Learning (SHDL) Network”, filed on Sep. 10, 2019, which is incorporated by reference herein in its entirety and for all purposes.FIELD OF THE INVENTION[0002]The present invention relates to a Disguised Face identification (DFI) system and method for identifying individuals with disguised faces in uncontrolled environments / uncontrolled scenarios. More particularly, the invention relates to a ScatterNet Hybrid Deep Learning (SHDL) Network for identifying individuals with disguised faces. The ScatterNet Hybrid Deep Learning (SHDL) Network detects facial landmarks or key points from the disguised face; then these facial landmarks or key points are utilized to form a unique face-specific signature; and this unique signature...

AI Technical Summary

Benefits of technology

[0022]One advantage of the present invention is an interactive Disguised Face Identification (DFI) with Facial Key Points using the ScatterNet Hybrid Deep Learning (SHDL) Network.

[0023]One advantage of the present invention is the system identifies individuals with the disguised faces including a wide variety of altered physical attributes on the face or wearing numerous disguises such as but not limited to wearing a wig, changing hairstyle or hair color, wearing eyeglasses, removing or growing beards, wearing scarves, wearing caps, wearing mask etc.

[0024]Another advantage of the present invention is the system identifies or recognizes the disguised faces of the individual at different orientations and distances.

[0025]Another advantage of the present invention is the system identifies or recognizes the individuals with the disguised faces in uncontrolled environments / scenarios.

[0026]Another advantage of the present invention is the system identifies or recognizes multiple individuals with different disguises in uncontrolled scenarios. This is possible as the system is trained on a large dataset that contains faces with varied disguises, covering different backgrounds and under varied illuminations. This allows the system to perform robust face recognition in the presence of different disguises and background variations.

[0027]Another advantage of the present invention is the deployed system can also be evolved by the user as he / she can add new faces to the database, by simply clicking the face on the monitor screen (interactive). The ScatterNet Hybrid Deep Learning (SHDL) Network detects the facial landmarks for these newly added faces without the need for extensive complete dataset training and can recognize these newly added faces in uncontrolled environments when seen thereafter.

Problems solved by technology

Therefore, extracting features from the complete face may not be the best solution as features from the disguises are also included in the extracted features.

Similarly, feature corruption will be present if the features are extracted from around the landmarks as several landmarks may be hidden by the disguises.

These factors may degrade the performance of the system.

However, most face recognition applications such as criminal identification, tracking school attendance, facilitate secure financial transactions, recognize VIPs at sporting events, and require the face recognition technology to work in uncontrolled environments.

The deployed model can't be changed or adapted to learn additional new faces after deployment.

This is a major drawback as the addition of new faces requires the complete retraining of the model on the complete dataset which requires extensive effort.

Images